INNOVATIVE STRUCTURE OF A PLAIN ELECTROMAGNETIC COMPONENT

Description

[0001] The present invention relates to the field of planar magnetic components, such as inductors, coupled inductors, and transformers. More specifically, the invention concerns an innovative planar transformer structure.

[0002] Currently, almost all switched-mode power supplies incorporate magnetic components. These components can be purchased off-the-shelf and added to the design or developed in-house. The invention relates to this second possibility and, in particular, to a category of magnetic components called planar magnetics. The main idea behind this technology is to integrate the component windings within the PCB. Planar magnetic components are a solution for power integration. These components are notably made using flattened magnetic cores (ferrite) and windings fabricated on the printed circuit board (PCB). The advantages of these planar magnetic components are numerous: they allow for better integration of the component into the design, the reproducibility of the component's electrical characteristics is increased, they allow for custom component design and therefore, optimization for the application.

[0003] There figure 1 This schematic diagram represents an example of the implementation of a planar magnetic component 5 according to the state of the art. This component 5 consists of an electrical circuit 6, made up of one or more windings 7, themselves composed of one or more turns (7-1, 7-2, 7-3, 7-4). These windings are designed to produce a magnetic field. This field can be used for energy storage (inductance) or for energy transfer (transformer). The component 5 includes a ferromagnetic core 8, which channels the magnetic field. This is referred to as the magnetic circuit. This core 8 can be made of various materials depending on the target application (power / frequency / price / size / performance). The core 8 may include an air gap 9, a small air space in the circuit, extending parallel to the plane of the circuit.

[0004] The flow of current in an electrical circuit generates losses in the same way as the flow of a magnetic field in a magnetic circuit. The losses in the two elements are respectively called copper losses and iron losses. These losses are interdependent. It is therefore desirable to optimize the dimensions of each element according to the application in order to maximize the overall performance. US2002 / 070735A1 discloses a transformer having the characteristics of the preamble of claim 1. J. Schäfer et al.: "Novel Highly Efficient / Compact Automotive PCB Winding Inductor Based on the Compensating Air-Gap Fringing Field Concept", IEEE Transactions on Power Electronics, Vol. 35, No. 9, 24.01.2020, pp. 9619-9633, describes a printed multilayer inductor having connection vias at the center of the winding around the perimeter of a central opening in the printed circuit board.

[0005] The invention aims to overcome all or part of the problems mentioned above by proposing a transformer comprising an innovative electromagnetic component structure that optimizes transformer performance by minimizing losses, improving PCB (Printed Circuit Board) integration by limiting vias around components, limiting parasitic inductances and improving couplings.

[0006] To this end, the invention relates to a transformer comprising: a primary circuit comprising a primary winding of N1 turns of an electrical conductor, the primary winding extending from a primary input terminal to a primary output terminal; and a secondary circuit comprising a secondary winding of N2 turns of an electrical conductor, the secondary winding extending from a secondary input terminal to a secondary output terminal, N1 and N2 each being an integer greater than or equal to 1; a printed circuit extending along a first plane, and comprising a plurality of layers superimposed with each other and forming an opening through the first plane about a first axis and defining a perimeter; a ferromagnetic core, disposed around the primary and secondary windings, comprising a central part disposed in the opening;a plurality of vias arranged at the center of the primary and secondary windings around the perimeter of the opening, and extending through the layers, each along an axis parallel to the first axis, the plurality of vias being configured to interconnect the plurality of layers; in that the N1 turns and the N2 turns of the electrical conductor are arranged each on one of the plurality of layers, according to any alternation between the N1 turns and the N2 turns, each of the N1 turns and the N2 turns winding, from a first via of the plurality of vias, partially around the plurality of vias forming an arc of a circle per layer, towards a second via of the plurality of vias; and in that the arc of a circle of one layer is distinctly oriented with respect to the arcs of the circles of the other layers and of distinct orientation from the arcs of the circles of the other layers.

[0007] Advantageously, the ferromagnetic core includes an air gap extending along a second axis substantially perpendicular to the first plane.

[0008] Advantageously, the input terminals are superimposed on the output terminals along a third axis substantially perpendicular to the foreground.

[0009] Advantageously, at least one among the plurality of layers is a shielding plane, preferably a ground plane.

[0010] The invention will be better understood and other advantages will become apparent upon reading the detailed description of an embodiment given by way of example, a description illustrated by the accompanying drawing in which: there Fig.1 schematically represents an example of the implementation of a planar magnetic component according to the state of the art; the Fig.2 schematically represents an example of the arrangement, around the central vias, of the primary and secondary windings of a transformer according to the invention; Fig.3 schematically represents an example of vias arranged at the center of the primary windings of an inductor according to the invention; the[ Fig.4 schematically represents an example of the implementation of the windings of a transformer according to the invention; the Fig.5 schematically represents the variation in current density according to a traditional air gap arrangement and an air gap arrangement according to the invention; the Fig.6 schematically represents the induction between the conductors according to the alternating turns of the primary and secondary windings; the Fig.7 schematically represents the homogenization of current density in the input and output terminals of the primary and secondary windings arranged according to an embodiment of the invention; the Fig.8 schematically represents an example of the implementation of a shielding layer in a transformer according to the invention; the Fig.9 schematically represents a classic electrical diagram of a synchronous rectifier; the Fig.10 schematically represents the optimization of output terminals for synchronous rectification according to the invention.

[0011] For clarity, the same elements will have the same reference points in the different figures. For better visibility and improved understanding, the elements are not always shown to scale.

[0012] There figure 1 schematically represents an example of the implementation of a planar magnetic component 5 according to the state of the art and has already been described in the introduction.

[0013] There figure 2 Figure 10 schematically represents a transformer according to the invention, with an example of the arrangement of the primary and secondary windings around the central vias. In this figure, the main elements of the transformer are represented by layers (normally superimposed). Note that this is an illustration; the number of layers is indicated only as a non-limiting example. A person skilled in the art understands that this number of layers may be greater or less than that shown in the figure. The transformer 10 comprises a primary circuit 11 including a primary winding 12 of N1 turns of a conductive electrical wire, the primary winding 12 extending from a primary input terminal 13 to a primary output terminal 14.The transformer 10 includes a secondary circuit 21 comprising a secondary winding 22 of N2 turns of a conductive electric wire, the secondary winding 22 extending from an input secondary terminal 23 to an output secondary terminal 24 (N1 and N2 each being an integer greater than or equal to 1). The transformer 10 comprises a printed circuit 15 (decomposed in the figure into several layers) extending along a first plane 16, and comprising a plurality of layers 17-1, 17-2, 17-3, 17-4, 17-5, 17-6, 17-7 superimposed on each other and forming an opening 18 through the first plane 16 around a first axis Z1 and defining a perimeter 19. The transformer 10 comprises a ferromagnetic core 25 (not shown in this figure, but intended to be inserted into the opening 18, and arranged around the primary winding 12 and secondary winding 22, comprising a central part 26 arranged in the opening 18).The transformer 10 comprises a plurality of vias 27 arranged at the center of the primary 12 and secondary 22 windings on the periphery 19 of the opening 18, and extending through the layers 17-1, 17-2, 17-3, 17-4, 17-5, 17-6, 17-7, each along an axis parallel to the first axis Z1, the plurality of vias 27 being configured to interconnect the plurality of layers 17-1, 17-2, 17-3, 17-4, 17-5, 17-6, 17-7.

[0014] According to the invention, the N1 turns and N2 turns of the conducting electrical wire are each arranged on one of the plurality of layers, in any alternation between the N1 turns and the N2 turns. In other words, there is one turn (either of the primary winding or of the secondary winding) per layer. This "any alternation" means that, in their superposition, a turn of the primary winding can be superimposed on a turn of either the primary or secondary winding. All combinations of superposition between primary and secondary windings are conceivable. Each of the N1 turns and N2 turns winds, from a first via of the plurality of vias 27, partially around the plurality of vias 27 forming an arc of a circle 28 per layer, towards a second via of the plurality of vias 27. In other words, per layer, the turn of the winding (primary or secondary) is not a complete turn, the turn does not make the 360° around the opening 18.Thus, some vias per layer are not surrounded by the tower. The central arrangement of the vias provides great flexibility in the positioning of the layers, which can be interleaved with respect to each other, and therefore in the positioning of the towers of the primary and secondary windings.

[0015] Furthermore, the circular arc 28 of one layer is distinctly oriented with respect to the circular arcs 28 of the other layers and has a distinct orientation from the circular arcs of the other layers. A turn, at the level of the perimeter 19 of the opening 18, can be considered as having a first end and a second end near the perimeter. The first and second ends are spaced a certain number of vias apart. This spacing between the first and second ends is found on each of the layers, and the respective spacings of the layers are not superimposed.

[0016] The transformer according to the invention allows for better integration and facilitates the implementation of shielding to further limit the impact of leakage flux near the air gap. Minimizing induction at the interconnections reduces losses. All these aspects and advantages of the invention are detailed below.

[0017] There figure 3 Figure (b) schematically represents an example of vias 27 arranged at the center of the winding 12 of an inductor 10 according to the invention. In this illustration, it should be considered that the diagram (b) is repeated six times and offset each time. This results in an inductor with 7 turns of a conductive electrical wire (therefore N1 = 3), implemented in 8 layers (again, only three layers are shown for better readability of the figure). The winding 12 extends from the primary input terminal 13 to the primary output terminal 14.

[0018] The use of vias at the center of the magnetic component simplifies the construction of the various windings. This allows for the reproduction of an elementary winding on each layer (b) to achieve the desired winding configuration. Only one turn is required per PCB layer. The transition between the different layers is achieved via the central vias 27. One or more vias can be used for this purpose, depending on the desired current in the windings and the size of the core 25 (and its central portion 26).

[0019] This one-turn-per-layer configuration goes against established practice. Indeed, usually in power electronics, the number of turns is spread across a single layer (as shown in the diagram). figure 1 Considering one turn per layer here necessitates a large number of PCB layers if a high number of turns are desired. Conversely, placing the vias in the center, around the perimeter of the opening, reduces layer-by-layer access resistances and frees up space around the component's periphery, allowing for better integration.

[0020] There figure 4 This schematically represents an example of the implementation of the primary winding 12 and secondary winding 22 of a transformer 10 according to the invention. More specifically, the output winding is incorporated within the ring of central vias 27. In order to interleave the primary and secondary windings, the vias enabling the interconnections between the layers 17 are also interleaved. The turns of the secondary winding can each be inserted between two turns of the primary winding and / or between one turn of the primary winding and one turn of the secondary winding. This configuration is advantageous for a transformer since it allows for better integration and facilitates the implementation of shielding to further limit the impact of proximity effects (and only in the case where the component has an air gap). Minimizing the induction at the interconnections reduces losses.

[0021] The figure 5 The diagram schematically represents the variation in current density according to a traditional air gap arrangement (left side of the figure) and an air gap arrangement according to the invention (right side of the figure). This representation is based on an illustration from the Schafer 2018 publication, "Optimal Design of Highly Efficient and Highly Compact PCB Winding Inductors." According to the invention, the ferromagnetic core 25 includes an air gap 29 extending along a second axis Z2 substantially perpendicular to the first plane 16. The use of a vertical air gap 29 is made possible by machining existing cores or raw material. Commercially available planar cores most often have an air gap located on the central leg, which causes the field to radiate in a direction parallel to the planar windings (see illustration on the left).The configuration on the left of the figure represents a copper conductor in the center subjected to leakage fields emanating from the two air gaps in the magnetic core. The current densities are concentrated at the edges of the conductor, which reduces the efficiency of the solution. More precisely, in a traditional (horizontal) air gap arrangement, the magnetic field propagates through the core. At the air gap, the field lines radiate around it, and these field lines tend to concentrate the currents flowing in the conductor outwards, to such an extent that the current flows only on the outside, where the field lines concentrate it. In other words, only a small portion of the conductor is actually used. In the configuration on the right of the figure, corresponding to the invention, the leakage fields arrive perpendicularly to the conductor, which allows for a reduction in current density and therefore in losses.More specifically, in a vertical configuration, the field radiates perpendicularly (see illustration on the right), which reduces the effects of proximity to the core and therefore decreases the current concentration at the ends of the electrical circuit. The currents are concentrated at the surface, and the entire conductor is used. This results in a positive impact on radiation. Thus, the resistance of the winding is reduced.

[0022] There figure 6 This diagram schematically represents the magnetic induction between the conductors according to the alternating turns of the primary and secondary windings. The lower part of the figure shows the conductors in a planar transformer. The layers labeled P1 represent the primary conductors, while the layers labeled S1 represent the secondary conductors. On the left side of the figure, the turns of the primary and secondary windings are arranged alternately; the choice of the alternating configuration is facilitated according to the invention. On the right side of the figure, the turns of the primary and secondary windings follow one another without any alternation between the primary and secondary windings. The theoretical magnetic induction profile (H) is given in the same diagram. The magnetic induction between the conductors increases the concentration of currents in them, which in turn increases losses.It can be seen that without alternating current, the maximum induction obtained is greater than the maximum induction obtained in the case of a transformer according to the invention (with alternating turns). This generates significant conduction losses between the two central layers (P1 and S1), which have a much higher resistance.

[0023] There figure 7 This schematic representation illustrates the homogenization of current density in the input and output terminals of the primary and secondary windings arranged according to an embodiment of the invention. This representation is based on an illustration from the Schafer 2018 publication, "Optimal Design of Highly Efficient and Highly Compact PCB Winding Inductors." In this embodiment of the invention, the input terminals 13, 23 are superimposed on the output terminals 14, 24 along a third axis Z3 substantially perpendicular to the first plane 16, as can be seen in the upper right of the figure. This avoids field concentration phenomena between the two planes. With the terminals positioned in two different parallel planes, the current is more evenly distributed throughout the plane and not just concentrated in the middle of a single plane.The lower part of the figure represents the results of a finite element simulation of the current density with adjacent terminals (on the left of the figure) and superimposed terminals according to the invention (on the right of the figure).

[0024] The results show that interlacing the conductors reduces the induction between them and therefore the current concentrations. A vertical arrangement of the terminals homogenizes the current densities and thus reduces losses in the terminals.

[0025] There figure 8 Figure 16 schematically represents a cross-sectional view, in a plane perpendicular to the first plane, of an example of the implementation of a shielding layer in a transformer according to the invention. In one embodiment of a transformer of the invention, at least one of the plurality of layers 27 is a shielding plane 31, preferably a ground plane. The shielding plane concentrates the eddy currents that generate losses. Thus, thanks to the shielding plane, these losses are generated within the shielding plane and no longer in the windings. The aim is to limit the total losses. The equivalent resistance of the circuit depends on the various resistances within the circuit. With a shielding plane, this resistance is reduced.

[0026] The shielding plane 31 is most often a ground plane. The leakage field creates an induced current (eddy current) in this plane, which generates losses. The distance from the shield to the air gap, the shield thickness, and the distance from the shield to the conductor depend on the power involved, the operating frequency (and signal shape), and the desired performance relative to the component integration.

[0027] In general, implementing the solution is beneficial if it reduces total losses. In the specific use case of a resonant converter, reducing the equivalent resistance of the conductors is a factor to consider. Limiting this resistance facilitates primary resonance and therefore smooth switching. In this particular case, it will also be necessary to consider the benefit of this operation on the magnetic sizing.

[0028] The invention improves the overall performance of a planar magnetic component through a set of features with numerous advantages: The central arrangement of vias facilitates the interconnection of different layers, particularly in the case of transformers with vias for the primary and secondary windings, and provides flexibility in the choice of interleaving (i.e., interleaved with each other). The presence of a shielding plane combined with a vertical air gap limits the effect of leakage flux on the conductors. This advantage is even more valuable in a resonant configuration. The optimization of the output terminals improves synchronous rectification. This advantage is especially beneficial in converters with high output current and high operating frequencies that require the use of one or more GaN transistors.

[0029] There figure 9 This schematic diagram represents a classic electrical circuit of a synchronous rectifier. On the left of this figure is the transformer (ideal coupler), Rs represents the parasitic series resistance of the secondary winding and the routing, QR the synchronous rectifier transistor, and DQR and CQR the parasitic components associated with this transistor. Cout and Rout represent the output capacitance of the converter and the load, respectively.

[0030] In the example we will examine, only two planes allow for the secondary winding. It is possible to imagine a different configuration to optimize performance (more copper in the secondary means less loss).

[0031] There figure 10Figure schematically represents the optimization of the output terminals for synchronous rectification according to the invention. As described previously, the winding can be implemented using every other via group. By optimizing the transformer terminals, it is possible to improve the secondary integration in order to minimize losses in the synchronous rectification. Generally, a voltage drop is applied between the primary and secondary windings. This results in a lower secondary voltage than primary voltage, which also means higher secondary currents. It is desirable to minimize the resistance on the secondary terminals to optimize performance. In the figure, the current path to the output is minimized.

[0032] This improvement leads to a reduction in RS resistance and parasitic inductances on the secondary side. Furthermore, it allows for an easier increase in the number of transistors in the synchronous rectification circuit, which further reduces losses.

[0033] Finally, it is thus possible to place the drivers as close as possible to the transistors, a critical point for GaN transistors for example.

[0034] It can be noted that the optimization of the various parameters of the magnetic components discussed above is adaptable to most converter configurations.

[0035] Thus, the invention comprises several technical features, which can be combined with each other, and whose technical effects are listed below: Use of vias positioned at the center of the planar core (near the central section). This configuration allows for easier distribution of the different windings without compromising external integration within the component. This arrangement also simplifies layer interleaving. Use of an air gap machined on top of the magnetic core. Unlike a horizontal air gap arrangement, a vertical arrangement orthogonal to the windings limits proximity effects and thus reduces copper losses, especially at high frequencies (>500 kHz). Interleaving / superposition of terminals on a vertical plane. Interleaving reduces induction and therefore high current concentrations. The vertical arrangement allows for the use of the full cross-section of the planar conductors, thus reducing AC resistance. Use of shielding planes.Located as close as possible to the air gap, they limit the effects of proximity on the conductors. The vertical air gap arrangement combined with the shielding minimizes the effects of the air gap on the conductors. Terminals are optimized for the integration of GaN transistors. Since synchronous rectification operates at high frequency and high current, it is necessary to limit parasitic inductances and resistances in the secondary winding. An interleaved and optimized secondary winding arrangement increases the performance of this type of system.

[0036] It will more generally become apparent to the person skilled in the art that various modifications can be made to the methods of implementation described above, in light of the teaching which has just been disclosed to him.

Claims

1. Transformer (10) comprising: - a primary circuit (11) comprising a primary winding (12) of N1 rotations of an electrical conductor, the primary winding (12) extending from a primary input terminal (13) to a primary output terminal (14); and - a secondary circuit (21) comprising a secondary winding (22) of N2 rotations of an electrical conductor, the secondary winding (22) extending from a secondary input terminal (23) to a secondary output terminal (24), N1 and N2 each being an integer greater than or equal to 1; - a printed circuit (15) extending along a first plane (16), and comprising a plurality of layers (17-1, 17-2, 17-3, 17-4, 17-5, 17-6, 17-7) superimposed to one another, and forming an opening (18) through the first plane (16) about a first axis (Z1) and defining a perimeter (19); - a ferromagnetic core (25), disposed around the primary (12) and secondary (22) windings, comprising a central part (26) disposed in the opening (18); - a plurality of vias (27) disposed at the centre of the primary (12) and secondary (22) windings over the perimeter (19) of the opening (18), and extending through the layers (17-1, 17-2, 17-3, 17-4, 17-5, 17-6, 17-7), each about an axis parallel to the first axis (Z1), the plurality of vias (27) being configured to interconnect the plurality of layers; in that the N1 rotations and the N2 rotations of the electrical conductor are each disposed on one of the plurality of layers, according to any alternance between the N1 rotations and the N2 rotations, characterised in that each of the N1 rotations and the N2 rotations being wound, from a first via of the plurality of vias (27), partially around the plurality of vias (27) forming one circular arc (28) per layer, to a second via of the plurality of vias (27); and in that the circular arc (28) of a layer is distinctly oriented with respect to the circular arcs (28) of the other layers, and of distinct orientation of the circular arcs of the other layers.

2. Transformer according to claim 1, wherein the ferromagnetic core (25) comprises an air gap (29) extending about a second axis (Z2) substantially perpendicular to the first plane (16).

3. Transformer according to claim 1 or 2, wherein the input terminals (13, 23) are superimposed to the output terminals (14, 24) about a third axis (Z3) substantially perpendicular to the first plane (16).

4. Transformer according to any one of claims 1 to 3, wherein at least one from among the plurality of layers (27) is a shielding plane (31), preferably a ground plane.